Examining Three Levels of Social Integration and Health in Minorities: A Bioecological Perspective

Abstract

Background. Understanding and dismantling health disparities remain one of the most essential imperatives for our nation. The complexity of interacting sociodemographic and structural factors affecting health is difficult to quantify. Thus, sophisticated approaches which take into account not only an individual but their dynamic, social environments are necessary for understanding resiliency and strengths in these populations (Thornton et al., 2016). In this dissertation, I adopted a developmental perspective (e.g. Bioecological Theory) that may guide scientists when considering several interacting, sociocultural environmental factors at once. Social integration is a powerful force in an individual's life. Although 'social integration' may have many names (perceived social support, closeness of social ties, diverse social networks, etc.), broadly, it is accepted that one's social life has a profound impact on their corresponding health and mortality through various behavioral and physiological pathways (Holt-Lunstad, Smith, & Layton, 2010; Robles & Kiecolt-Glaser, 2003; Symister & Friend, 2003). Thus, any comprehensive study on health should include a broader investigation of socio-environmental variables including measures of social integration and broader community culture and resources. Historically, however, investigations focused on social integration, physical health, and the association between the two have traditionally underrepresented minority individuals (Heiat, Gross, & Krumholz, 2002; Hussain-Gambles, Atkin, & Leese, 2004; Murthy, Krumholz, & Gross, 2004). Research Aim and Questions. Adopting a social-ecological systems approach, the focus of my dissertation is to examine social integration at three levels (the spousal/partner relationship, immediate family-level dynamics, and neighborhood and community level factors) and the association of these levels of social integration with the health of minority individuals (Bronfenbrenner & Evans, 2000; Bronfenbrenner & Morris, 1998, 2007). I addressed this aim through three, separate research questions (RQ)/Chapters: RQ1/Chapter II) The individual and spousal/partner relationship, What are the associations between marital status and social support with health quality of life (HQoL), and mortality in post-menopausal, Hispanic women? RQ2/ Chapter III) Immediate family-level dynamics, Does synchrony of emotional arousal in diverse families facing breast cancer predict depressive mood and coping style in breast cancer patients? RQ3/ Chapter IV) Neighborhood and community level factors, Do social ties or social support mediate the ethnic density effect for Hispanics and other racial-ethnic groups? Methods and Data Sources. Three distinct data sources were examined within this body of work. In Chapter II, I utilized data from the Women's Health Initiative, Observational Study (Anderson, et al., 1998). In this analysis, I examined the relation between marital status and two outcomes: mortality and health quality of life. I also assessed whether the relations between marital status and these outcomes were attenuated or moderated by social support and language acculturation. In Chapter III, I utilized data collected for The Family Coping and Breast Cancer Project which recruited patients during the years of 1991 – 1993 (Weihs et al., 2005). In this analysis, I examined emotion arousal synchrony among family members (patient and spouse, and patient and child) using cross recurrence quantification analysis. I then examined whether the relations between emotion arousal synchrony and two breast cancer patient outcomes (coping style and depressive mood) are moderated by patient perceived family environment ratings or race. Lastly, in Chapter IV, I analyzed data from Wave 2 of the National Social Life, Health, and Aging Project (O’Muircheartaigh, English, Pedlow, & Kwok, 2014). In this analysis, I examined whether the Hispanic ethnic density effect was mediated by either social network variables (size, closeness of ties, and diversity) or social support and whether these associations were moderated by Hispanic ethnicity. Results. In the first study (Chapter II) I found that marital status was a significant predictor of mortality for older Hispanic women. Specifically, widowed women had significantly higher mortality risk when compared to their married counterparts. This relation, however, was not present after controlling for social support and language acculturation. In regard to health quality of life, marital status was associated with physical functioning, with widowed women reporting significantly worse physical functioning three years after baseline when compared with their married counterparts. This study suggests that widowed, Hispanic women may be at risk for poor health and this may be facilitated through social support and language acculturation. In the second study (Chapter III) I found that the relation between emotional arousal synchrony and patients' depressive mood in families facing breast cancer was moderated by family environment, specifically conflictual environments. I also found that the relation between emotional arousal synchrony and coping style was moderated by family environment, specifically for cohesiveness ratings. Moderation by race was not found. These results suggest that emotion arousal synchrony are family dynamics that may have differing implications depending on the family environment. In the third study (Chapter IV) I found two Hispanic ethnic density effects, but they were not mediated by social network variables. However, they were mediated by social support, but not in the direction I hypothesized. Social support was an inconsistent mediator of the relation between ethnic density and depressive symptoms and a suppressor of the relation between ethnic density and morbidity. These results suggest that for all racial/ethnic groups, ethnically dense neighborhoods do not beget higher social ties and social support. Conclusions. Overall, findings in this dissertation suggest that varying levels of the social-ecological environment are associated with health in minorities in various ways. A common phenomenon that arose in all three analyses were questions about how covariates or mediators influenced the association of the main variable of interest and its relation to different health outcomes. Broadly, this may be a common issue for social scientists interested in utilizing social variables from different social-ecological levels (micro-system, exosystem, etc.), which corroborates my previous statement regarding the complexity of social environments.